Prioritizing Venture Assumptions by Uncertainty and Impact

Hello. Last lesson turned your opportunity thesis into a map of assumptions about desirability, viability, feasibility, trust, and regulatory risk. The map makes the uncertainties visible; it does not tell you which one deserves attention first.

Today you’ll use an assumption-prioritization matrix to rank those claims by two questions: How much does this venture depend on the claim being true? and How much relevant evidence do you have for it? The result should guide your next discovery decisions without requiring you to commit to an app—or treat early customer conversations as proof.

Judge impact by what would change if you were wrong

An assumption is high impact when finding it false would undermine an important part of the venture. The consequence may be abandoning a proposed segment, changing the value proposition, or reconsidering whether you can responsibly deliver a particular concept. An assumption is lower impact when you could drop a feature without substantially changing the opportunity.

Consider the illustrative thesis from last lesson: people who describe difficult-to-manage flow and work shifts with restricted breaks want fewer interruptions and less worry about bleeding through. The following claims would not all have the same impact:

• Existing preparations still leave consequential disruption during those shifts. If this is false for the proposed segment, the opportunity thesis needs substantial revision.
• People would find a private record of flow experiences useful. If this is false, a recording concept may need to change, but another form of support might still address the need.
• People would value optional educational tips alongside that record. If this is false, the tips could be removed.

Impact is therefore relative to the venture you are considering, not a judgment that one person’s experience matters more than another’s. Write down the consequence of being wrong before assigning a position on the matrix. “We would need a different feature” and “we would need a different customer problem” are materially different outcomes.

A conditional claim needs special care. Suppose you are considering an automated alert that purports to identify medically abnormal bleeding. Whether the associated claims and functionality can be responsibly supported could become a high-impact question if that alert is essential to your concept. It is not automatically a top venture assumption while you remain open to concepts that make no such judgment. Keep the condition attached to the claim rather than letting an imagined feature determine the priority of the solution-neutral opportunity.

Judge uncertainty by evidence for the specific claim

It is tempting to place an assumption on a “known–unknown” scale according to how confident you feel. A better prompt is: What have you observed that bears directly on this statement?

There is a difference between evidence that a problem exists and evidence for a proposed response to it. Published accounts of disruptive heavy flow can make your opportunity plausible, but they do not establish how often disruption persists among people working restricted-break shifts. An interviewee’s account of a recent difficult shift would bear more directly on that segment-specific claim. Even then, their account would not establish that they would adopt a private-record tool or pay for one.

Evidence also varies with the behavior you are trying to predict. Descriptions of past workarounds can illuminate a current need. A favorable reaction to a hypothetical product is a weaker signal of future sustained use. Observed use is more direct evidence of use than stated interest, though a small test still leaves questions about how widely its findings apply. Avoid a universal ladder that treats one method as strongest for every claim; ask how closely the observation matches the people, context, and behavior named in this assumption.

In this segment of Testing Business Ideas: Assumptions Mapping Webinar, David J. Bland demonstrates how to compare assumptions by importance and evidence. Notice his distinction between hearing someone express interest and observing a more consequential behavior.

For your customer conversations, retain an evidence note beside each assumption: what was observed, whose experience it reflects, when it happened, and what the observation does not establish. Your qualitative-research experience is useful here: a vivid account can reveal a mechanism or a neglected context without serving as an estimate for everyone in a proposed market.

Contrary evidence matters, too. If relevant participants consistently describe their existing preparations as sufficient, the answer is not to leave the assumption labeled “uncertain” indefinitely. Record what challenges it and reconsider the thesis. The matrix prioritizes learning; it is not a device for protecting favored beliefs.

Place claims on the matrix, then rank within it

Put potential impact if false on the vertical axis, increasing upward. Put lack of relevant supporting evidence on the horizontal axis, increasing to the right. Thus, the upper-right area contains claims that matter greatly but remain weakly supported.

This is a continuum, not four objective boxes. A useful first pass is to give each claim a provisional placement and a short explanation:

Here is a hypothetical first pass, not a characterization of what your conversations have established. Each claim is separate so evidence for one cannot silently stand in for another.

The table illustrates an important distinction: A has some indirect evidence but may outrank C, even though C has less evidence, because A tests the foundation of the proposed opportunity. You might provisionally rank A first, B second, and C third. A asks whether this particular segment has the unresolved problem. B asks whether people seek help beyond their current preparations. C matters greatly if you pursue private recording, but it does not decide whether the broader opportunity exists.

That order is a reasoned judgment, not a calculation. If your conversations already contain several detailed, recent accounts from people in the proposed segment supporting A, move A leftward and reconsider the order. If your concept becomes inseparable from collecting intimate data, C’s impact rises. Do not multiply arbitrary “impact” and “uncertainty” scores and mistake the result for measured risk. Relative positions and explicit reasons are more useful than false precision.

The five lenses from last lesson can be shown with labels or colors on the same matrix; they are not five separate ranking systems. A and B concern desirability, D viability, E feasibility, and C trust. Keep regulatory questions visible as well. For example, “Our intended functions and claims can be pursued on a pathway we can responsibly support” is too broad until you specify the functions and claims. Once a particular health-related feature becomes central, its review may be consequential even if customer evidence is your immediate discovery focus.

A priority for investigation is not permission to proceed in the meantime. If a contemplated test would collect real sensitive health information, unresolved privacy and security obligations are a reason to pause that collection and obtain appropriate guidance—not simply another sticky note to investigate after launch. Likewise, do not put an unreviewed medical claim in front of prospective users because a matrix ranks some desirability question higher. The matrix allocates learning effort; it does not override safeguards.

Make the ranking usable—and revise it

A compact assumption record can prevent a lively mapping discussion from becoming an unexplained set of sticky-note positions. For each leading claim, retain:

1. The claim and its condition: for example, “If we choose private recording…”
2. The consequence if false: what would you change or stop?
3. Current evidence: direct, indirect, contradictory, or absent, with its source and limits.
4. Its relative rank: why it sits above or below the next claim.

If you work with collaborators, have each person place claims provisionally before discussing them. A disagreement often reveals the most useful information: perhaps one person knows of relevant customer accounts, while another has assumed those accounts demonstrate product interest. Neither the most confident voice nor the person who proposed the concept should determine the evidence position.

Keep the wording reasonably discrete, but do not spend the whole session perfecting every statement. Map first; sharpen the highest-priority claims enough to guide the next stage. In this short segment, Bland explains why the map should change as evidence arrives.

For example, later interviews might move A left because accounts from the proposed segment consistently describe consequential disruption. They could also reveal that the most pressing unmet need occurs in a different context, reducing A’s relevance to your revised thesis. A prototype might provide evidence about E’s effort burden without answering D’s payment question. Keep each movement tied to what was actually learned.

Takeaways

The assumption-prioritization matrix compares the consequence of being wrong with the strength of evidence for the precise claim. Its upper-right area deserves attention, but the strongest ranking also considers dependencies: an unresolved problem thesis may matter more than an untested feature of one possible solution. Conditional concepts, trust concerns, and regulatory or privacy safeguards should remain explicit rather than being flattened into a single score.

You now have a way to turn last lesson’s assumption inventory into a defensible shortlist, with reasons you can revisit. Next, you’ll use that shortlist to define research questions, methods, and decision thresholds for an evidence-based discovery plan.